PS: PLoS announces PLoS Clinical Trials, new open accessjournal

["Peter Banks" <pbanks@diabetes.org>]

An interesting postscript to this topic appeared today in JAMA. The same
Cleveland Clinic researcher who raised red flags about Vioxx today raised
very serious concerns about a new diabetes drug, Pargluva, that had
already received an approvability letter from FDA. See
<http://jama.ama-assn.org/misc/294.20.joc50147.pdf>

I take back the suggestion that we need a new journal to critically
analyze clinical trials. JAMA is doing a fine job already (and made the
new study freely available, perhaps because of its immediate significance
for clinical care.)

Peter Banks
Acting Vice President for Publications/Publisher
American Diabetes Association
Email: pbanks@diabetes.org

>>> pbanks@diabetes.org 10/19/05 8:52 PM >>>

I have trouble seeing how this journal addresses the problem of selective
reporting of positive clinical trials.  PLoS would have us believe that
selective reporting is the fault of traditional journals (as apparently
are many of the world's problems): "Traditional medical journals publish
only the highest profile clinical trials (typically positive trials),
partly because the journals must attract revenues from subscriptions and
selling reprints."

Journals are not the root problem of selective reporting, which has
multiple causes--the non-submission of negative trials by funders, the
failure to vigorously follow up on safety problems suggested in published
trials, and the weakness of regulatory agencies like the FDA in the drug
approval process. Cleveland Clinic cardiologist Eric Topol has published
an excellent and instructive analysis of the failure to examine safety
problems evident in various published and unpublished Vioxx trials
(http://ccjm.org/PDFFILES/Karha12_04.pdf). The existence of PLoS Clinical
Trials would have done little to expose problems inherent in drugs like
Vioxx. In fact, Topol published a 2001 analysis in that most traditional
of journals, JAMA, that suggested a substantial caridovascular risk from
Vioxx--an analysis that failed to attract attention until the APPROVe
trial led to the drug's withdrawl in 2004.

The ICMJE statement on clinical trial registries tries to get at the
problem of selective reporting by ensuring that journal editors and
reviewers, as well as clinicians and the general public, know of at least
the existence of all studies in a given field. The registries do not
contain the results, of course, but the non-publication of trial findings
may raise an index of suspicion.

There may be a need for a clinical trial journal to expose the Vioxx
problems in the making, but PLoS Clinical Trials is probably not it.

What is really needed is a publication that will publish the work of
investigative researchers who will take the time to rigorously analyze all
available published data, then dig into the registries to find the
unreported findings that may expose the problems not apparent in the
published data.

Peter Banks
Acting Vice President for Publications/Publisher
American Diabetes Association
Email: pbanks@diabetes.org